Combinatorial model of ligand-receptor binding

Overview

Combinatorial model of ligand-receptor binding

The binding of ligands to receptors is the starting point for many import signal pathways within a cell, but in contrast to the specificity of the processes that follow such bindings, the bindings themselves are often non-specific. Namely, a single type of ligand can often bind to multiple receptors beyond the single receptor to which it binds optimally. This property of ligand-receptor binding naturally leads to a simple question:

If a collection of ligands can bind non-specifically to a collection of receptors, but each ligand type has a specific receptor to which it binds most strongly, under what thermal conditions will all ligands bind to their optimal sites?


Depiction of various ligand types binding optimally and sub-optimally to receptors

In this repository, we collect all the simulations that helped us explore this question in the associated paper. In particular, to provide a conceptual handle on the features of optimal and sub-optimal bindings of ligands, we considered an analogous model of colors binding to a grid.


Partially correct and completely correct binding for the image

In the same way ligands could have certain receptors to which they bind optimally (even though such ligands could bind to many others), each colored square has a certain correct location in the image grid but could exist anywhere on the grid. We have the correct locations form a simple image so that when simulating the system it is clear by eye whether the system has settled into its completely correct configuration. In all of the notebooks in this repository, we use this system of grid assembly as a toy model to outline the properties of our ligand-receptor binding model.

Reproducing figures and tables

Each notebook reproduces a figure in the paper.

Simulation Scheme

For these simulations, we needed to define a microstate, the probability of transitions between microstates, and the types of transitions between microstates.

Microstate Definition

A microstate of our system was defined by two lists: one representing the collection of unbound particles, and the other representing particles bound to their various binding sites. The particles themselves were denoted by unique strings and came in multiple copies according to the system parameters. For example, a system with R = 3 types of particles with n1 = 2, n2 = 3, and n3 = 1 could have a microstate defined by unbound_particles = [A2, A2, A3] and bound_particles = [A1, −, A2, −, A1, −] where “−” in the bound list stands for an empty binding site.

Since the number of optimally bound particles was an important observable for the system, we also needed to define the optimal binding configuration for the microstates. Such an optimal configuration was chosen at the start of the simulation and was defined as a microstate with no unbound particles and all the bound particles in a particular order. For example, using the previous example, we might define the optimal binding configuration as optimal_bound_config = [A1, A1, A2, A2, A2, A3], in which case the number of optimally bound particles of each type in bound_particles = [A1,−,A2,−,A1,−] is m1 = 1, m2 = 1, and m3 = 0. The number of bound particles of each type is k_1 = 2, k_2 = 1, and k_3 = 0. We note that the order of the elements in unbound_particles is not physically important, but, since the number of optimally bound particles is an important observable, the order of the elements in bound_particles is physically important.

For these simulations, the energy of a microstate with k[i] bound particles of type i and m[i] optimally bound particles of type i was defined as

E(k, m) = Sum^R_i (m[i] log delta[i] + k[i] log gamma[i])

where k=[k1,k2,...,,kR] and m=[m1,m2,...,mR], gamma[i] is the binding affinity, and delta[i] is the optimal binding affinity of particle of type i. For transitioning between microstates, we allowed for three different transition types: Particle binding to a site; particle unbinding from a site; permutation of two particles in two different binding sites. Particle binding and unbinding both occur in real physical systems, but permutation of particle positions is unphysical. This latter transition type was included to ensure an efficient-in-time sampling of the state space. (Note: For simulations of equilibrium systems it is valid to include physically unrealistic transition types as long as the associated transition probabilities obey detailed balance.)

Transition Probability

At each time step, we randomly selected one of the three transition types with (equal probability for each type), then randomly selected the final proposed microstate given the initial microstate, and finally computed the probability that said proposal was accepted. By the Metropolis Hastings algorithm, the probability that the transition is accepted is given by

prob(init → fin) = min{1, exp(- β(Efin −Einit))*π(fin → init)/π(init → fin) }

where Einit is the energy of the initial microstate state and Efin is the energy of the final microstate. The quantity π(init → fin) is the probability of randomly proposing the final microstate state given the initial microstate state and π(fin → init) is defined similarly. The ratio π(fin → init)/π(init → fin) varied for each transition type. Below we give examples of these transitions along with the value of this ratio in each case. In the following, Nf and Nb represent the number of free particles and the number of bound particles, respectively, before the transition.

Types of Transitions

  • Particle Binding to Site: One particle was randomly chosen from the unbound_particles list and placed in a randomly chosen empty site in the bound_particles list. π(fin → init)/π(init → fin) = Nf^2/(Nb +1).

Example: unbound_particles = [A2, A2, A3] and bound_particles = [A1, −, A2, −, A1, −]unbound_particles = [A2, A3] and bound_particles = [A1, A2, A2, −, A1, −]; π(fin → init)/π(init → fin) = 9/4

  • Particle Unbinding from Site: One particle was randomly chosen from the bound_particles list and placed in the unbound_particles list. π(fin → init)/π(init → fin) = Nb/(Nf + 1)^2.

Example: unbound_particles = [A2, A2, A3] and bound_particles = [A1, −, A2, −, A1, −]unbound_particles = [A2, A2, A3, A2] and bound_particles = [A1, −, −, −, A1, −]; π(fin → init)/π(init → fin) = 3/16

  • Particle Permutation: Two randomly selected particles in the bound_particles list switched positions. π(fin → init)/π(init → fin) = 1.

Example: unbound_particles = [A2, A2, A3] and bound_particles = [A1, −, A2, −, A1, −]unbound_particles = [A2, A2, A3] and bound_particles = [A2, −, A1, −, A1, −]; π(fin → init)/π(init → fin) = 1

For impossible transitions (e.g., particle binding when there are no free particles) the probability for accepting the transition was set to zero. At each temperature, the simulation was run for anywhere from 10,000 to 30,000 time steps (depending on convergence properties), of which the last 2.5% of steps were used to compute ensemble averages of ⟨k⟩ and ⟨m⟩. These simulations were repeated five times, and each point in Fig. 6b, Fig. 7b, Fig. 8b, and Fig. 9 in the paper represents the average ⟨k⟩ and ⟨m⟩ over these five runs.

References

[1] Mobolaji Williams. "Combinatorial model of ligand-receptor binding." 2022. [http://arxiv.org/abs/2201.09471]


@article{williams2022comb,
  title={Combinatorial model of ligand-receptor binding},
  author={Williams, Mobolaji},
  journal={arXiv preprint arXiv:2201.09471},
  year={2022}
}
Owner
Mobolaji Williams
Mobolaji Williams
Official pytorch implementation of "DSPoint: Dual-scale Point Cloud Recognition with High-frequency Fusion"

DSPoint Official pytorch implementation of "DSPoint: Dual-scale Point Cloud Recognition with High-frequency Fusion" Coming soon, as soon as I finish a

Ziyao Zeng 14 Feb 26, 2022
A collection of educational notebooks on multi-view geometry and computer vision.

Multiview notebooks This is a collection of educational notebooks on multi-view geometry and computer vision. Subjects covered in these notebooks incl

Max 65 Dec 09, 2022
[ICML 2021] Break-It-Fix-It: Learning to Repair Programs from Unlabeled Data

Break-It-Fix-It: Learning to Repair Programs from Unlabeled Data This repo provides the source code & data of our paper: Break-It-Fix-It: Unsupervised

Michihiro Yasunaga 86 Nov 30, 2022
Road Crack Detection Using Deep Learning Methods

Road-Crack-Detection-Using-Deep-Learning-Methods This is my Diploma Thesis ¨Road Crack Detection Using Deep Learning Methods¨ under the supervision of

Aggelos Katsaliros 3 May 03, 2022
Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation

Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation [Arxiv] [Video] Evaluation code for Unrestricted Facial Geometry Reconstr

Matan Sela 242 Dec 30, 2022
Collection of generative models in Pytorch version.

pytorch-generative-model-collections Original : [Tensorflow version] Pytorch implementation of various GANs. This repository was re-implemented with r

Hyeonwoo Kang 2.4k Dec 31, 2022
Implementation of Vaswani, Ashish, et al. "Attention is all you need."

Attention Is All You Need Paper Implementation This is my from-scratch implementation of the original transformer architecture from the following pape

Brando Koch 195 Dec 30, 2022
Data and Code for ACL 2021 Paper "Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning"

Introduction Code and data for ACL 2021 Paper "Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning". We cons

Pan Lu 81 Dec 27, 2022
This is the code of "Multi-view Contrastive Graph Clustering" in NeurlPS 2021.

MCGC Description This is the code of "Multi-view Contrastive Graph Clustering" in NeurlPS 2021. Datasets Results ACM DBLP IMDB Amazon photos Amazon co

31 Nov 14, 2022
1st Place Solution to ECCV-TAO-2020: Detect and Represent Any Object for Tracking

Instead, two models for appearance modeling are included, together with the open-source BAGS model and the full set of code for inference. With this code, you can achieve around 79 Oct 08, 2022

Improving Calibration for Long-Tailed Recognition (CVPR2021)

MiSLAS Improving Calibration for Long-Tailed Recognition Authors: Zhisheng Zhong, Jiequan Cui, Shu Liu, Jiaya Jia [arXiv] [slide] [BibTeX] Introductio

DV Lab 116 Dec 20, 2022
RL and distillation in CARLA using a factorized world model

World on Rails Learning to drive from a world on rails Dian Chen, Vladlen Koltun, Philipp Krähenbühl, arXiv techical report (arXiv 2105.00636) This re

Dian Chen 131 Dec 16, 2022
Visualizer using audio and semantic analysis to explore BigGAN (Brock et al., 2018) latent space.

BigGAN Audio Visualizer Description This visualizer explores BigGAN (Brock et al., 2018) latent space by using pitch/tempo of an audio file to generat

Rush Kapoor 2 Nov 21, 2022
Source code related to the article submitted to the International Conference on Computational Science ICCS 2022 in London

POTHER: Patch-Voted Deep Learning-based Chest X-ray Bias Analysis for COVID-19 Detection Source code related to the article submitted to the Internati

Tomasz Szczepański 1 Apr 29, 2022
PCGNN - Procedural Content Generation with NEAT and Novelty

PCGNN - Procedural Content Generation with NEAT and Novelty Generation Approach — Metrics — Paper — Poster — Examples PCGNN - Procedural Content Gener

Michael Beukman 8 Dec 10, 2022
The official implementation of paper "Finding the Task-Optimal Low-Bit Sub-Distribution in Deep Neural Networks" (IJCV under review).

DGMS This is the code of the paper "Finding the Task-Optimal Low-Bit Sub-Distribution in Deep Neural Networks". Installation Our code works with Pytho

Runpei Dong 3 Aug 28, 2022
Pytorch library for seismic data augmentation

Pytorch library for seismic data augmentation

Artemii Novoselov 27 Nov 22, 2022
Implementation of NÜWA, state of the art attention network for text to video synthesis, in Pytorch

NÜWA - Pytorch (wip) Implementation of NÜWA, state of the art attention network for text to video synthesis, in Pytorch. This repository will be popul

Phil Wang 463 Dec 28, 2022
Graph InfoClust: Leveraging cluster-level node information for unsupervised graph representation learning

Graph-InfoClust-GIC [PAKDD 2021] PAKDD'21 version Graph InfoClust: Maximizing Coarse-Grain Mutual Information in Graphs Preprint version Graph InfoClu

Costas Mavromatis 21 Dec 03, 2022
Plug-n-Play Reinforcement Learning in Python with OpenAI Gym and JAX

coax is built on top of JAX, but it doesn't have an explicit dependence on the jax python package. The reason is that your version of jaxlib will depend on your CUDA version.

128 Dec 27, 2022